Re: [CAD_CAM_EDM_DRO] Re: Batteries in Power Supply

I worked in Medical Imaging service for several years, and the practice
of having the cells from the same mfg. & lot # was SOP for the portables
using Lead acid cells.
Bill Higdon

sparkness2001 wrote:

> Ive worked on telephone switch battery supplies. when replacing lead
> acid batteries they must all be from the same mfg and lot# period in
> order to charge and prevent problems between the 24 cells.
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Carl Mikkelsen, Oasis"
> <Carl.Mikkelsen@o...> wrote:
>
>>Just some thoughts ...
>>
>>It was suggested that the capacitors could be eliminated if a
>
> battery was
>
>>floated across the power supply output. This started me thinking
>
> about the
>
>>battery problem.
>>
>>I would be concerned about the high ripple current through the
>
> battery, and
>
>>constant switching between charge and discharge cycles (120 times
>
> per second).
>
>>I would not build my own system to work this way, and would put
>
> significant
>
>>filter capacitance on the output.
>>
>>I would also used lead-acid batteries, because they are rated for
>>"floating" on the mains.
>>I would use a regulated output supply, set to the proper float
>
> voltage for
>
>>the batteries I was using.
>>I would make the power supply be able to gracefully and safely
>
> handle the
>
>>over current condition, and allow power to come from the batteries
>
> (current
>
>>fold back regulation).
>>
>>I would oversize the batteries, making the batteries peak discharge
>
> current
>
>>be five times the expected peak servo load. [This may be too
>
> conservative,
>
>>and I could change my mind after seeing the battery prices. :) ]
>>
>>My father would have built this by taking a 12.6 V filament
>
> transformer,
>
>>running it through a selenium bridge rectifier, across some caps,
>
> through a
>
>>light bulb, across the battery, and off to the load.
>>
>>Since we like to use higher voltages as CNC supplies, I'd be
>
> concerned
>
>>about cell-to-cell variance when placing a large number of cells
>
> (lead-acid
>
>>or otherwise) in series. Each lead-acid cell has a voltage of
>
> about 2
>
>>volts, so you must stack 30 cells to reach 60 volts. That allows a
>
> lot of
>
>>cells for temperature variances, chemistry changes, and other nasty
>
> things
>
>>to build up and cause problems.
>>
>>Some one of you who is a battery engineer will probably point out
>
> the
>
>>fallacies in what I've said. Even if some of my concerns are
>
> valid, there
>
>>may be rational ways to engineer around them.
>>
>>It is an area where a little engineering up-front (temp sensors,
>
> current
>
>>monitoring, maybe a little closed-loop charging control) could save
>
> a lot
>
>>of grief later.
>>
>>I'm sure it is possible to make it work. The phone company had 48V
>>lead-acid batteries floating across the line supply for decades,
>
> and maybe
>
>>still does. It's been quite a few years since I've been inside a
>
> phone
>
>>exchange building.
>>
>>-- Carl
>>
>>
>>At 03:27 PM 1/21/2003 -0500, you wrote:
>>
>>>So would this mean that you could eliminate the filter capacitor
>>
> from the
>
>>>power supply leaving only the transformer and rectifier?
>>>
>>>John Guenther
>>>'Ye Olde Pen Maker'
>>>Sterling, Virginia
>>>
>>>
>>>>-----Original Message-----
>>>>From: Mariss Freimanis <mariss92705@y...>
>>>>[mailto:mariss92705@y...]
>>>>Sent: Tuesday, January 21, 2003 14:14 PM
>>>>To: CAD_CAM_EDM_DRO@yahoogroups.com
>>>>Subject: [CAD_CAM_EDM_DRO] Re: Batteries in Power Supply
>>>>
>>>>
>>>>Bernard,
>>>>
>>>>I posted on this topic about a year or so ago. The main
>>>
> advantage
>
>>>>mentioned was servomotors usually have a large ratio between
>>>
> their
>
>>>>stall current and continuous rated current (5:1 or more).
>>>>
>>>>This means the power supply has to be 5 or more times bigger
>>>
> than
>
>>>>needed just to supply very short duty cycle accel/decel current
>>>>pulses to the motor.
>>>>
>>>>The idea was to use low AH batteries to supply this current;
>>>
> they
>
>>>>would recharge between accel/decel cycles. The battery would
>>>
> connect
>
>>>>to the power supply "+" terminal via a rectifier and it's
>>>
> voltage
>
>>>>would be a little less than the power supply's voltage. Under
>>>
> load,
>
>>>>the supply voltage would sag, forward biasing the rctifier from
>>>
> the
>
>>>>battery, which would now supply the pulse current.
>>>>
>>>>A motor stall current 5 times the max rated current implies a
>>>
> 4% duty
>
>>>>cycle. Theoretically a 1AH battery could supply 20A for 3
>>>
> minutes,
>
>>>>way longer than you would need for accel/decel (1 or 2 seconds).
>>>>
>>>>Now the supply could be sized for the continous current only
>>>
> plus
>
>>>>what would be needed to restore the battery charge (5 * 4% =
>>>
> 20% or
>
>>>>1.2 times the max rated motor current).
>>>>
>>>>Mariss
>>>>
>>>>--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Bernard R
>>>><bwjarandall@c...>" <bwjarandall@c...> wrote:
>>>>
>>>>>In view of the recent interest in power supplies has anyone
>>>>
> thought
>
>>>>>about using batteries in place of capacitors in some of the
>>>>
> larger
>
>>>>>power supplies?
>>>>>
>>>>>If you used 5 12Volt batteries and 1 6Volt that would give a
>>>>
>>>>nominal
>>>>
>>>>>66 volt, with a fully charged cell rising to 2.4 volt that
>>>>
> would
>
>>>>give
>>>>
>>>>>an additional 13.2Volt (33 cells @ .2V) for a total of 79.2
>>>>
> volt.
>
>>>>>In a recent post Les Watts reports routinely seeing 60 amp
>>>>
> peaks on
>
>>>>>rapid starts, stops and sharp curves, I would think batteries
>>>>
> could
>
>>>>>cope with these conditions better than conventional caps.
>>>>>
>>>>>I'm not sure if there would be any significant price
>>>>
> difference
>
>>>>>between caps and batteries, a 100Volt 15,000uF cap is almost
>>>>
> $30
>
>>>>and
>>>>
>>>>>for a reasonably stiff supply you probably need at least 3 in
>>>>>parallel. Without doing the math, I suspect you need more
>>>>
> than 3
>
>>>>caps
>>>>
>>>>>to cope with 60 Amp peaks. 12V 4.5 A/Hour batteries are about
>>>>
> $17
>
>>>>and
>>>>
>>>>>6V 4.2 A/H $9 for a total of $94. (12V 12 AH cost $$35)
>>>>>
>>>>>I'm not sure how effectively batteries kill spikes, I would
>>>>
> have
>
>>>>>thought much better than caps, they should absorb overvoltage
>>>>
> from
>
>>>>>motors quite effectively.
>>>>>
>>>>>I can see difficulties in disconnecting the batteries from the
>>>>
>>>>drives
>>>>
>>>>>at power off, but this should be offset by being able to use
>>>>
>>>>smaller
>>>>
>>>>>transformers.
>>>>>
>>>>>Just a thought.
>>>>>
>>>>>Bernard